Skip to content

FRM Exam Part II · Correlation Basics: Definitions, Applications, and Terminology

Financial Correlation Risk and Correlation Swaps Explained

Updated 11 October 2026 · Fact-checked

Financial correlation risk is the risk that the strength of co-movement between assets or risk factors changes and causes losses. A correlation swap exchanges realized correlation for a fixed strike. Payoff to the fixed-rate payer = notional × (realized correlation − strike). Check the sign, then multiply by notional.

Understand Financial Correlation Risk and Correlation Swaps

Financial correlation risk is the risk of loss because the relationship between two or more variables changes. The variables can be asset prices, default events, rates or volatilities. A position can look well hedged or diversified today and stop being so when correlations move.

It helps to separate two ideas. Static correlation describes co-movement at a point in time or over a fixed sample, for example the correlation of two stocks' returns over the last year. It is a single number and does not model how it changes. Dynamic correlation describes how co-movement evolves over time. Examples are mean reversion, jumps, or correlation linked to volatility or market state. A model that uses a static input for a long-dated product ignores this risk.

Correlation risk shows up in many places. In equity portfolios, diversification fails when stock correlations rise in a sell-off. In credit, default correlation drives multi-name products such as CDO tranches. In quanto and cross-asset options, the price depends on the correlation between the asset and the exchange rate. In basket and rainbow options, higher correlation raises the basket's volatility. Correlation also tends to rise in stressed markets, so risk models calibrated on calm periods can understate losses.

A correlation swap lets you trade realized correlation directly. One party pays a fixed correlation (the strike) and receives the realized correlation over the life of the swap, multiplied by a notional. The other party takes the opposite side. The realized figure is usually the average of pairwise correlations across a basket of assets, measured on a set schedule such as daily returns. The buyer of correlation gains if realized correlation ends above the strike.

Similar instruments include variance swaps on single stocks and an index. Dispersion trades combine them: a trader who sells index variance and buys single-stock variance is short correlation. Correlation swaps are OTC contracts, so they carry counterparty risk and are not cleanly hedgeable.

Key formulas to remember

Correlation swap payoff (fixed-rate payer, long correlation)
Payoff = N × (ρ_realized − K)
N is notional per unit of correlation, K is the strike. The short side receives the negative of this amount.
Average realized pairwise correlation
ρ_realized = [2 ÷ (n² − n)] × Σ(i<j) ρ(i,j)
For n assets there are n(n − 1) ÷ 2 distinct pairs. This is an equally weighted average.
Basket variance (equal weights, equal volatility σ, common correlation ρ)
σ_basket² = σ² × [1/n + (1 − 1/n) × ρ]
Higher ρ raises basket volatility. This is why index volatility relative to single-stock volatility reveals implied correlation.
Two-asset portfolio variance
σ_p² = w₁²σ₁² + w₂²σ₂² + 2w₁w₂ρσ₁σ₂
Portfolio risk rises with ρ, other things equal.
Correlation range
−1 ≤ ρ ≤ +1
A strike or realized value outside this range is not possible for Pearson correlation.

How to solve Financial Correlation Risk and Correlation Swaps questions

Use this method for any question on correlation risk or correlation swap payoffs.

  1. 1Identify what is being asked: a concept (static vs dynamic, where correlation risk arises) or a calculation (swap payoff, average correlation, basket variance).
  2. 2For a swap, find the strike K, the realized correlation, the notional, and which side the question asks about.
  3. 3Decide the sign. The long-correlation (fixed-rate payer) side gains when realized correlation exceeds K.
  4. 4If realized correlation must be computed from pairs, count the pairs as n(n − 1) ÷ 2 and average them.
  5. 5Compute N × (ρ_realized − K) and keep the currency, such as USD, in the answer.
  6. 6For concept questions, ask whether the risk is that correlation changes over time (dynamic) or the use of one fixed input (static), and which direction stress moves it.
  7. 7Check the answer against intuition: a rise in correlation hurts diversified long-only portfolios and helps long-correlation positions.

Quickest way: Sign-first swap shortcut

When to use it: Use it for MCQs on correlation swap payoffs or on who benefits from a correlation move.

  1. Write down realized minus strike for the long-correlation side.
  2. If positive, the long side gains and the short side pays; if negative, reverse it.
  3. Multiply by the notional and scale units carefully, since correlation of 0.10 is 10 points.
  4. Eliminate options with the wrong sign first, then check the magnitude.

Common mistakes in Financial Correlation Risk and Correlation Swaps

  • Getting the payoff sign wrong

    Students forget that the fixed-rate payer is long correlation and profits when realized exceeds the strike.

    Fix: Always compute realized − strike for the long side, then flip for the short side.

  • Counting pairs as n² or n instead of n(n − 1) ÷ 2

    It is easy to count both ρ(i,j) and ρ(j,i), or to include self-correlation.

    Fix: For 4 assets the number of distinct pairs is 6, not 16 or 12.

  • Treating static and dynamic correlation as the same

    Both are called correlation and both can be estimated from data.

    Fix: Static is a point-in-time or fixed-sample value. Dynamic describes how it changes, for example mean reversion or stress-related spikes.

  • Assuming diversification always helps in stress

    Students use calm-period correlations as if they are permanent.

    Fix: State that correlations tend to rise in stressed markets, so diversification benefit can shrink when it is most needed.

  • Saying correlation measures any dependence

    Pearson correlation is a common default.

    Fix: Pearson captures linear dependence only. Nonlinear or tail dependence can exist with low Pearson correlation.

Worked examples

Example 1

A USD 2 million-per-correlation-point-of-1.00 correlation swap (notional N = USD 2,000,000 per 1.00 of correlation) has a strike of 0.40. Realized correlation over the term is 0.55. What does the fixed-rate payer (long correlation) receive?

Show the solution
  1. Payoff = N × (ρ_realized − K).
  2. ρ_realized − K = 0.55 − 0.40 = 0.15.
  3. Payoff = 2,000,000 × 0.15 = USD 300,000.

Answer: The long-correlation party receives USD 300,000; the short party pays it.

Example 2

A basket has 5 assets. The sum of all distinct pairwise correlations is 4.0. What is the average realized pairwise correlation, and what is the payoff to the short-correlation side on a USD 1,000,000 notional with strike 0.50?

Show the solution
  1. Number of distinct pairs = 5 × 4 ÷ 2 = 10.
  2. Average correlation = 4.0 ÷ 10 = 0.40.
  3. Long-side payoff = 1,000,000 × (0.40 − 0.50) = −USD 100,000.
  4. The short side takes the opposite sign: +USD 100,000.

Answer: Average correlation is 0.40, and the short-correlation side receives USD 100,000.

Exam tips

  • Read which side the question asks about before calculating; sign errors are the most common lost mark.
  • Be ready to link correlation to products: basket options, quanto options, CDO tranches and dispersion trades.
  • Know that rising stress correlation reduces diversification and can invalidate models calibrated on calm data.
  • For static vs dynamic, give one clear example of each rather than a general definition.

Practice questions from Correlation Basics: Definitions, Applications, and Terminology

Financial Correlation Risk and Correlation Swaps: frequently asked questions

What is financial correlation risk?

It is the risk of loss from changes in the co-movement of assets, defaults or risk factors. A hedge or diversification benefit can fail when correlation shifts, especially in stressed markets.

How does a correlation swap work?

One party pays a fixed correlation strike and receives realized correlation on a notional. At maturity the net amount is notional × (realized − strike), paid to whichever side is owed.

What is the difference between static and dynamic correlation?

Static correlation is a single value over a fixed period or point in time. Dynamic correlation models how the relationship changes over time, including mean reversion and jumps.

Who benefits when correlation rises?

The long-correlation side of a swap gains. Diversified long-only portfolios and dispersion sellers tend to lose, since diversification weakens.